A new drone surveillance system can spot when someone in a crowd is acting violently. It uses artificial intelligence and is going to be tested at a university festival in India later this year.
The system assesses the way each person in a crowd is standing via two cameras on the drone. Currently it can recognise punching, stabbing, shooting, kicking and strangling.
However, once the technology goes on sale, whoever buys it will be able to input their own definition of a violent act, says Amarjot Singh听at the University of Cambridge, who built the system with two colleagues based at science institutes in India.
Advertisement
To teach the system what a violent act looks like, the user moves computer-generated stick people on a screen into the different motions they want the drone to recognise.
In initial tests, the system was 94 per cent accurate at identifying violent acts. However, as more people appear in frame, it became less accurate, falling to 79 per cent accuracy when looking at 10 people simultaneously.
The drone surveillance system is going to be tested in October at Technozion, a science and technology festival Andra Pradesh, India. 鈥淭he festival is attended by around 3,000 to 4,000 people and is extremely packed,鈥� says Singh.
Ethical concerns
However, there are concerns about the technology. 鈥淲hy is this being trialled at college festivals in India rather than at Glastonbury or Coachella?鈥� says Malavika Jayaram, a cyber law researcher at Harvard, and a fellow of the Centre for Internet and Society in India. She suggests it may be because it is easier legally and socially to get approval.
Violence can鈥檛 just be described in such a basic set of motions, and technology like this is likely to miss an important aspect of identifying crime 鈥� intent, says Jayaram.
There are also issues around misuse,听says Gemma Galdon Clavell, who works on the ethics of technology at the University of Barcelona. 鈥淲hat if this tech is used by non-democratic regimes to identify dissidents? Or by gangs to identify enemies?鈥�
The system is an example of imperfect science attempting to enact the fantasy of a perfect society, says Jayaram. 鈥淓ven if the accuracy rates were higher, is this the kind of society we want to build?鈥�
Singh and his colleagues certainly think so. 鈥淎I can help develop powerful surveillance systems which can assist in identifying pernicious individuals which will make the society a safer place,鈥� says Singh. 鈥淭herefore, I think it is a good thing and is also necessary.鈥�
Reference:听
